Recently, large platform operators running
dual mode business, under which a platform runs a marketplace and sells its own
products in that marketplace, have attracted intense scrutiny from both
academics and policymakers. One of the concerns for such vertical integration
is that a platform’s usage of non-public data generated by business activities
of third-party sellers might dampen the entry and innovation of third-party
sellers. The EU and the US have, in response, discussed restricting such dual
mode practices of large platforms by various legal measures. Some would place
special restrictions on the platforms’ usage of marketplace data in creating
goods to enter their own marketplace. Despite these rapid changes in the
business environment, the growth of platform brands, and the intense legal
discussions on platform’s running dual mode businesses, our fundamental
understanding of this marketplace and how proposed restrictions would affect
innovation of third-party sellers is still limited, which is one of the main
determinants of welfare.
On June 28, 2022, Professor Erik Madsen from New York University joined us in
the Luohan Webinar to shed some light on this important question using a simple
but innovative theoretical model. The model consists of three actors: a
platform, an entrepreneur (third-party seller), and a regulator. Platform hosts
a marketplace for entrepreneur’s product and at the same time can introduce its
own private-label products into the marketplace. The product introduced by the
platform is going to compete with the entrepreneur’s product in the
marketplace. The entrepreneur can develop a new product to sell on marketplace
and is assumed to have no option to sell directly to consumers, a realistic one
for many small and medium sized sellers. In the model, the entry of an
entrepreneur with a newly invented product is regarded as a positive
innovation. Lastly, there is a regulator who can design and enforce a
regulation to restrict data usage of a platform.
Three things are worth mentioning to understand the model before discussing the main results of the model. First, the entry of an entrepreneur generates data on market demand state (denoted as α in the model) and this single variable captures all the information that allows the platform to judge whether this market (product) is worthwhile to enter or not. Second, the profit of the platform increases when it imitates the entrepreneur’s product, while the profit of the entrepreneur decreases. Lastly, consumer welfare improves as the level of innovation increases.
After explaining the basics of the model, Prof. Madsen compared three different forms of regulations on data usage (about α): no-regulation, total data-ban, and data patent. First, he compared the equilibrium innovation levels of entrepreneurs under no regulation and under a total data ban. Interestingly, his result shows that no-regulation does not always reduce innovation; a data ban increases innovation only when the product’s demand is “experimental.” Experiment products are those whose demand is dichotomous: either large success or failure. The platform under no-regulation can see the realization of the demand, and make profits by exploiting the entrepreneur’s experimental product whenever it hits the market. On the other hand, without getting access to the data, the platform enters an average market. As a result, data-ban increases innovation compared to the one under no-regulation. What is interesting is that data-ban decreases the innovation of “non-experimental (‘incremental’ in the paper)” products due to the opposite effect in the same mechanism: that the platform still enters products with more stable demand, and thus discourages the entrepreneur. Therefore, the impact on innovation from switching from no-regulation to data-ban is dependent upon whether the demand of a product is “experimental” or not.
The second main result showed how the innovation under data-patent compares the innovation under no-regulation and data-ban. Data-patent is a regulation that restricts the usage of data by the platform for certain period of time. Data-patent is different from regular patent because the platform can imitate even before the data-patent period expires, it just cannot use the data so its entry decision is not dependent upon the data (α in the model). Since the regulator can always achieve the same level of innovation using the data-patent, data-patent always outperforms data-ban in terms of the level of innovations by entrepreneur. Prof. Madsen further showed that data-patent can outperform even no-regulation by choosing the optimal the length of the patent, which depends on the degree to which the product is experimental. Lastly, Prof. Madsen also showed that data-patent always induces more innovation than divesture, which is one of the remedies suggested by some politicians and antitrust administrators. This is because, in the model, a platform operator earns platform fees for selling third-party goods and thus must constrain the sale of its own goods, while a divested firm only earns profit from selling its own good.
Recent discussions on platform vertical integration and data-appropriation sometimes tend to lead to extreme remedies such as total data-ban or divesture. This paper provides us with a new perspective to look at this issue of platform dual mode and points out a key underlying trade-off in there. It shows that platforms economics are complex and dependent on multiple variables. Moreover, the main results of the paper suggest that a regulator might induce more innovations by implementing data-regulation tailored to demand characteristics of distinct product categories.
About the Speaker: Erik Madsen is an Assistant Professor of Economics at the New York University. His research focuses on Microeconomic Theory, with a special focus on incentive contracting, dynamic games, and industrial organization. His research has been published in leading academic journals, including the American Economic Review and Review of Economics Studies.
Paper: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3832712
If you would like to give a presentation in a future webinar, contact our Senior Economist Dr. Wen Chen (wen.chen@luohanacademy.com). For other inquiries, please contact: events@luohanacademy.com.
